Home B101 B142L B202 B302 B324 M520 Mosaic Course updatesPresentation format
<http://cnfolio.com/NotesInventoryModels>
Electronics Manufacturing – B324

Inventory models




Economic order quantity (EOQ) is the classical inventory model for stock held on cycle with assumptions of known demand and production lead time. The objective is to select an order quantity that minimizes the marginal annual costs for holding inventory and placing orders.

Economic order quantity




Let D be the steady annual demand for a product and Q be the order size for instant re-stock of inventory.

The average inventory level is:
Q ÷ 2

Let H be the marginal annual cost of holding one unit of inventory. Then, the marginal annual cost for all products is:
H x ( Q ÷ 2 ) = ( H x Q ) ÷ 2


The average order frequency is:
D ÷ Q

Let A be the marginal annual cost of placing one order. Then, the marginal annual cost for all orders is:
A x ( D ÷ Q ) = ( A x D ) ÷ Q


The economic order quantity (EOQ) is the point at which the marginal holding cost equals the marginal ordering costs, which is calculated as:
Economic order quantity


The total annual cost of inventory is the sum of the marginal annual cost for holding inventory and placing orders:
( ( H x Q ) ÷ 2 ) + ( ( A x D ) ÷ Q )

The total annual cost of inventory (TC) is easier to calculate using the EOQ value:
TC = EOQ x H




The annual demand for a product is 2,000. The unit cost of each product is £100 and the storage cost per annum amounts to 20% of stock value. Each order costs £30 for processing. What is the optimal number of orders per year?

From the problem description, A is £30 and D is 2,000.

H is calculated as a percentage of the unit cost:
£100 x 0.20 = £20


The first step of the EOQ calculation is:
2 x £30 x 2,000 = £120,000

then,
£120,000 ÷ £20 = 6,000

and finally,
the square root of 6,000 is approximately 77


Using 77 as the Q value, the optimal number of orders per year is approximately:
2,000 ÷ 77 = 26




On average, 10 products are sold daily and the retailer is open 240 days per year. Each product costs £50 and the holding cost is 24% of the product cost. There is a cost of £9 per order. What is the total annual inventory cost at the optimal inventory level?

From the problem description, A is £9

D is calculated from the daily sales:
10 x 240 = 2,400


H is calculated as a percentage of the unit cost:
£50 x 0.24 = £12


The first step of the EOQ calculation is:
2 x £9 x 2,400 = £43,200

then,
£43,200 ÷ £12 = 3,600

and finally,
the square root of 3,600 is 60


Using 60 as the EOQ value, the total annual inventory cost is:
60 x £12 = £720




Approximately 68% of data values will be within one standard deviation of the average value.

68% of data within one standard deviation of the average value




Approximately 95% of data values will be within two standard deviation of the average value.

95% of data within two standard deviation of the average value




Approximately 99.73% of data values will be within three standard deviation of the average value (or 2,700 errors per million observations).

Approximately 99.9999998% of data values will be within six standard deviation of the average value (or 2 errors per million observations).

Six sigma of quality




Variability pooling depends on the statistical relationship where:
standard deviation of group of N items = ( square root of N ) x ( standard deviation of a single item )




A manufacturer has 20 retailer customers. Each customer orders a monthly average of 400 products with a standard deviation of 20 products. The manufacturer has committed to a 99.73% service level agreement for order fulfillment. What is the comparative efficiency between having 20 warehouses and 5 warehouses?

A service level agreement for 99.73% requires 3 standard deviations of variable inventory for each warehouse:
400 + ( 3 x 20 ) = 460 products

The base case of 20 warehouses would require the manufacturer to maintain a total monthly inventory level of:
460 x 20 = 9,200 products


The equivalent standard deviation for a group of 4 retailers being serviced by 1 warehouse (5 warehouses total) is:
( square root of 4 ) x ( 20 ) = 2 x 20 = 40 products

A service level agreement for 99.73% requires 3 standard deviations of variable inventory for each warehouse:
( 4 x 400 ) + ( 3 x 40 ) = 1,720 products

The case of 5 warehouses would require the manufacturer to maintain a total monthly inventory level of:
1,720 x 5 = 8,600 products


The efficiency of variability pooling is calculated as:
( 9,200 - 8,600 ) ÷ 9,200 = 600 ÷ 9,200 = 6.5% efficiency




Variability of risk between a manufacturer and a retailer customer can be reduced with a risk sharing contract such as product buy back:

The probability that a retailer customer will order enough to meet sales demand is expressed as a ratio:
( (retail price) - (wholesale price) ) ÷ ( (retail price) - (buy back price) )


The optimal ratio is obtained by setting wholesale price to the cost of manufacturing and the buy back price to zero:
( (retail price) - (manufacturing price) ) ÷ ( retail price )


The buy back price is calculated using the optimal ratio:
retail price - ( ((retail price) - (wholesale price)) ÷ (optimal ratio) )




The cost to manufacture one product is £20. The manufacturer sells at a wholesale price of £50 and its retail price is £100. What is the buy back price for optimal risk sharing between the manufacturer and the retailer customer?

The optimal ratio is calculated as:
( 100 - 20 ) ÷ 100 = 80 ÷ 100 = 80%


When the buy back price is £30, the ratio is low and would cause the retailer to keep inventory below the optimal quantity:
( 100 - 50 ) ÷ ( 100 - 30 ) = 50 ÷ 70 = 71%


When the buy back price is £40, the ratio is high and would cause the retailer to keep inventory above the optimal quantity:
( 100 - 50 ) ÷ ( 100 - 40 ) = 50 ÷ 60 = 83%


When the buy back price is £35, the ratio is low, but closer to optimal:
( 100 - 50 ) ÷ ( 100 - 35 ) = 50 ÷ 65 = 77%


When the buy back price is £37.50, the ratio is optimal:
( 100 - 50 ) ÷ ( 100 - 37.50 ) = 50 ÷ 62.50 = 80%


Which is the same as using the formula above:
100 - ( (100 - 50) ÷ 0.8 ) = 100 - ( 50 ÷ 0.8 ) = 100 - 62.50 = 37.50